WSREP status codes to be integrated into perror or printed or documented
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
MySQL patches by Codership |
New
|
Undecided
|
Unassigned | ||
Percona XtraDB Cluster moved to https://jira.percona.com/projects/PXC |
Confirmed
|
Wishlist
|
Raghavendra D Prabhu |
Bug Description
In instances like:
"wsrep::init() failed: %d, must shutdown",
an error code is printed without any explanation.
Following statuses defined in wsrep_api.h are returned:
typedef enum wsrep_status {
WSREP_OK = 0, //!< success
WSREP_WARNING, //!< minor warning, error logged
WSREP_
WSREP_TRX_FAIL, //!< transaction aborted, server can continue
WSREP_BF_ABORT, //!< trx was victim of brute force abort
WSREP_
WSREP_
WSREP_FATAL, //!< fatal error, server must abort
WSREP_
} wsrep_status_t;
So,
a) It needs to be integrated into perror framework (in which case
it needs to have error code with an offset as in errmsg-utf8.txt
)
b) Use strerror or equivalent to print the error in full at
WSREP_ERROR and other WSREP_* instances.
c) Document it somewhere.
a) and b) are preferrable.
Changed in percona-xtradb-cluster: | |
assignee: | nobody → Raghavendra D Prabhu (raghavendra-prabhu) |
status: | New → Triaged |
Changed in percona-xtradb-cluster: | |
milestone: | 5.5.33-23.7.6 → future-5.5 |
Changed in percona-xtradb-cluster: | |
status: | Triaged → Confirmed |
importance: | Undecided → Wishlist |
Also, seen in cases like
WSREP: applier thread exiting (code:0)
[Note] WSREP: applier thread exiting (code:5)
It is better to integrate it into perror framework.